Embed Size (px)
Citation preview
Improving
productivity for Australian pasture based dairy farming
A report for By Aubrey Pellett 2014 Nuffield Scholar March 2015 Nuffield Australia Project No 1416
Sponsored by:
ii
© 2013 Nuffield Australia. All rights reserved. This publication has been prepared in good faith on the basis of information available at the date of publication without any independent verification. Nuffield Australia does not guarantee or warrant the accuracy, reliability, completeness or currency of the information in this publication nor its usefulness in achieving any purpose. Readers are responsible for assessing the relevance and accuracy of the content of this publication. Nuffield Australia will not be liable for any loss, damage, cost or expense incurred or arising by reason of any person using or relying on the information in this publication. Products may be identified by proprietary or trade names to help readers identify particular types of products but this is not, and is not intended to be, an endorsement or recommendation of any product or manufacturer referred to. Other products may perform as well or better than those specifically referred to. This publication is copyright. However, Nuffield Australia encourages wide dissemination of its research, providing the organisation is clearly acknowledged. For any enquiries concerning reproduction or acknowledgement contact the Publications Manager on ph: (03) 54800755 . Scholar Contact Details Aubrey Pellett 45 Williams Rd Hill End Victoria 3825 Australia Phone: +61429667699 Email: [email protected]
In submitting this report, the Scholar has agreed to Nuffield Australia publishing this material in its edited form. NUFFIELD AUSTRALIA Contact Details Nuffield Australia Telephone: (03) 54800755 Facsimile: (03) 54800233 Mobile: 0412696076 Email: [email protected] PO Box 586 Moama NSW 2731
iii
Executive Summary
In the search for new technologies and management practices likely to improve productivity,
several themes were considered. These included how technology enables greater labour
efficiency, the growing and consumption of more feed per hectare, improving the feed
conversion efficiency of the dairy herd, and what does the emerging concept of precision
dairy mean?
The recent introduction of a robotic rotary milking system will significantly boost labour
productivity. High system capacity will suit Australia’s pasture based grazing systems
allowing one supervisor to milk cows in a batch fashion. Adoption of this technology will
require minimal farm or production system change, while providing physical relief and rich
information at the cow level. Dairy-‐specific algorithms are being developed that enable
generic robots to complete paddock related activities. Robotics tailored to the dairy farm
environment will facilitate continued farm size expansion by providing a solution to dairy
labour issues.
Ryegrass is likely to remain the base of the Australian dairy grazing system. The use of
genomic technology has the ability to speed up development of new ryegrasses with
improved yield and persistence. The provision of an industry cultivar selection tool in
development by Dairy Australia will guide farmers’ pasture selection decisions and provide
feedback to breeders by incentivising genuine cultivar improvement. Inoculation of the best
performing cultivars with novel and designer endophyte will boost both pasture and animal
performance. The introduction of fodder beet as a high yielding crop would benefit from the
development of Australian guidelines.
The development of precision dairy technology is in its infancy. Decisions based on real time
cow and paddock information can increase yield, reduce costly inputs, improve animal
iv
welfare, and enable farmers to focus on decision making rather than task completion. The
ability to collate data from multiple sources improves the accuracy and reliability of decision
support systems.
The capture and use of increasing quantities of information from the farm has the potential
to change how the supply chain is integrated. Greater ability to predict input demand and
milk supply will allow the farm to form value creating closer relationships and be more
responsive to price signals.
There is variation in feed conversion efficiency between lactating cows. The opportunity to
breed for improved feed conversion efficiency is now possible using genomic prediction. The
recommended approach is as a part of a broad, future looking objective, recognizing that in
the genomics era the speed of genetic gain is rapid. Using genomic information for individual
herds will allow customized management best suited to a herd’s predisposed genetic
response.
v
Contents
Executive Summary ........................................................................................................... iii
Contents ............................................................................................................................. v
Index of figures ................................................................................................................. vii
Foreword ......................................................................................................................... viii
Acknowledgments .............................................................................................................. x
Abbreviations ................................................................................................................... xii
Objectives ......................................................................................................................... 13
Chapter 1: Introduction .................................................................................................... 14
Chapter 2: Improving labour productivity ........................................................................ 17
Automated milking systems ................................................................................................. 18
Lely A4 milking robot ........................................................................................................ 18
De Laval voluntary milking system (VMS) ........................................................................ 19
De Laval automatic milking rotary (AMR) ........................................................................ 20
GEA DairyProQ ................................................................................................................. 22
Robotics ............................................................................................................................... 24
Fetching robot .................................................................................................................. 24
Heavy activity paddock robots ......................................................................................... 26
Conclusions .......................................................................................................................... 26
Chapter 3: Higher yields from home grown feed .............................................................. 27
Enhancing the productivity of ryegrass ............................................................................... 27
Genomics assisted breeding ............................................................................................. 27
Hybrid ryegrass ................................................................................................................ 28
vi
Transgenic high energy ryegrass ...................................................................................... 28
Endophyte selection ......................................................................................................... 29
Cultivar selection tools ........................................................................................................ 31
The pasture profit index -‐ Ireland .................................................................................... 31
The DairyNZ forage value index -‐ New Zealand ............................................................... 32
Fodder beet ......................................................................................................................... 34
Conclusions .......................................................................................................................... 35
Chapter 4: Precision dairy ................................................................................................. 37
Cow sensors ......................................................................................................................... 38
DairyMaster MooMonitor+ .............................................................................................. 39
Herd navigator ................................................................................................................. 39
Precision dairy in the paddock ............................................................................................. 40
Precision dairy challenges .................................................................................................... 41
Optimising the value chain .................................................................................................. 42
Conclusions .......................................................................................................................... 43
Chapter 5: Cow efficiency ................................................................................................. 45
What is cow efficiency? ....................................................................................................... 45
Gross efficiency ................................................................................................................ 45
Feed conversion efficiency or residual feed intake .......................................................... 46
Is the RFI difference significant? .......................................................................................... 46
Is it a feasible breeding objective? ....................................................................................... 47
Can RFI fit within a broader breeding objective? ................................................................ 48
Broader herd productivity ................................................................................................... 50
Genomic prediction .......................................................................................................... 50
vii
Precision genomic mating -‐ designer matings .................................................................. 50
Genomic guided farm management ................................................................................ 51
Conclusions .......................................................................................................................... 52
Recommendations ............................................................................................................ 53
References ........................................................................................................................ 55
Plain English Compendium Summary ................................................................................ 58
Index of figures
Photo 1: De Laval Automatic Milking Rotary at Camden, NSW, Australia. 21
Photo 2: First commercial DairyProQ installation in Germany. 23
Photo 3: Autonomous Robot trialled by FutureDairy in Camden, NSW, Australia. 24
Photo 4: Cultivar Field Trials, Ashburton, New Zealand. 30
Table 1: Calculation weightings in the Irish Pasture Profit Index 32
Table 2: New Zealand Forage Value Index Report Example. 33
Photo 5: Farmers Fodder Beet Sowing Date Trial, Windwhistle, New Zealand. 35
Photo 6: University of Kentucky Research Dairy Farm, Sensor Field Testing, United States. 38
Table 3: Milk solids production and feed intake for high and low efficient cows 46
Table 4: Actual difference in RFI when selected for RFI by genomic prediction 48
viii
Foreword
As a first generation dairy farmer I made the choice to start farming together with my
partner, Jacqui, in 2001. Our original motivation was to be self-‐employed, to be the prime
decision makers and to create lasting wealth. In the sometimes elusive pursuit of annual
profit I developed a recognition that a longer term focus on productivity was required. Even
a well managed dairy farm business has a profit influenced largely by fluctuating
international milk prices, rising input costs, and unpredictable climatic conditions. The desire
to achieve solid profits prompted a scan of the horizon to identify what promise the future
held.
I consider myself fortunate that my Nuffield scholarship has been sponsored by the Geoffrey
Gardiner Dairy Foundation. They have offered me every assistance and I appreciate the
informal discussions that helped guide my initial thinking. Conversations with some of the
dairy industry’s most experienced and talented people have greatly enhanced the Nuffield
experience.
In recent years I have been a Director of both Gippsland’s regional development program
GippsDairy and the Bonlac Supply Company which represents the interests of the farmers
who supply milk to Fonterra in Australia. Both these experiences are about benefiting
farmers and have given me a broader understanding of the research, development and
extension activities taking place in Australia. The combination of my own farm and these
industry experiences framed the judgments I made on where to look and how any new
technology would drive productivity in the Australian pasture based production system.
My Nuffield scholarship afforded me the opportunity to travel extensively in the second half
of 2014. My first trip as a part of an international Nuffield group provided a global view of
agricultural issues as we travelled through the Philippines, China, Canada, United States,
ix
Netherlands, France, and Ireland. China provided the chance to see first hand the impact on
food demand of increasing wealth, urbanization of a formerly rural population, and changing
consumer tastes. There can be no doubt that China is in the middle of a transition from small
scale tenant farmers to larger more mechanized farming businesses with the assistance of
central government planning. It will be interesting to see how they resolve the issue of land
ownership in the coming years as rural entrepreneurship is encouraged. A second trip to the
Netherlands, Germany, Ireland, Sweden and the United Kingdom enabled me to understand
how intensive systems and high labour costs promote development and uptake of
automated milking systems. My final trip through New Zealand and the United States
reinforced how a focus on pasture and the business of farming can deliver spectacular
results. There were many inspirational visits and meetings along the way.
Producing more from less is the challenge we need to overcome in order to remain
profitable and satisfy customers, while ensuring we pass on our natural resources in better
shape to future generations. In undertaking this Nuffield Scholarship and producing this
report I very much hope to paint a picture of what is possible in the future. If my own family
farming business and wider dairy industry become more competitive and profitable in the
future I will consider it to be a success.
x
Acknowledgments
My gratitude goes to Nuffield Australia for providing the opportunity and seeing potential in
me. My sponsor, the Geoffrey Gardiner Dairy Foundation, has been more than willing to
offer any assistance above and beyond the significant financial investment of a Nuffield
sponsor. Special thanks go to Mike, Mary, Barry and Aaron; your guidance has been
appreciated.
The scholars who joined me on the Global Focus Program made those six weeks of travel
one of the highlights of my Nuffield experience. I am lucky to have shared the experience
with Tania, Finola, Justine, Greg, Paul, Nigel, Nicky, and Steve.
To the many people around the world who kindly referred me to experts, organized
meetings, visits, and hosted me, your willingness to assist a Nuffield scholar has surprised,
delighted and humbled me.
I must acknowledge The Bonlac Supply Company Board for tolerating my absence and the
financial contribution to the experience I had with the Global Dairy Farmers Irish study tour.
A sincere thank you is due to the team at home ensuring the farm hardly noticed I was away;
in particular to Billy who stepped up to the challenge and taught me that the people around
you will strive to do their best.
Of course I can never thank my own family enough for the willing sacrifices they inevitably
had to make as I enjoyed the Nuffield experience. Thanks to my mother who was always
available to help Jacqui and the children, I never had to worry. To Anneka and Jackson, I am
confident that when you are older you will understand why Dad was away for what seemed
xi
like endless weeks. But most of all, none of this would have been possible without the
support of you, Jacqui. I always knew you would manage superbly in my absence and of
course you did.
xii
Abbreviations
VMS Voluntary milking system
AMR Automatic milking rotary
NSW New South Wales
GPS Global positioning system
CRC Cooperative Research Centre
DNA Deoxyribonucleic acid
GM Genetically modified
DM Dry matter
MJ Megajoule
LDH Lactate dehydrogenase
LIC Livestock Improvement Corporation
FCE Feed conversion efficiency
RFI Residual feed intake
CIDR Controlled internal drug release
13
Objectives
The objective of this report is to identify the new technologies, management practices and
other factors that are likely to improve dairy farm productivity in the near future,
specifically:
• What are the technologies that will be enablers of greater labour efficiency resulting
in more milk solids or cows managed per operator?
• How can a farm produce more quantity and quality of ‘home grown’ feed?
• Is it possible and practical to improve the herd’s efficiency of feed conversion?
• Does the growing ability to capture cow and farm data provide an opportunity to
make better decisions leading to precision dairy optimization?
Gaining an insight into yet to be released new technologies has proved difficult given the
commercial considerations of many innovations. The author experienced reluctance to
reveal new research and development beyond what is already released to the market place.
However, given the slow rate of adoption it is still possible to look at ‘new’ technology today
and assess its usefulness in addressing productivity barriers in the future.
14
Chapter 1: Introduction
The Australian dairy industry is a significant contributor to economic activity and as
Australia’s third largest rural industry is estimated to be worth $13 billion. Approximately $4
billion of this is received at farm gate by 6,314 farms that directly employ 43,000 people.
Total national milk production is 9.239 billion litres with 1.69 million cows producing on
average 5,471 litres. The industry is a major exporter responsible for seven per cent of world
dairy trade. Across Australia, the exportable surplus accounts for 38 per cent of total
production (Dairy Australia, 2013/14).
Over time Australian dairy farm businesses have become larger. Since 1983 the average herd
size of 90 cows has increased to an estimated 268 in 2014 (Dairy Australia, 2013/14). While
business challenges and opportunities have necessitated this trend it has been supported by
greater efficiencies resulting in more cows being managed per person, more feed being
produced per hectare farmed and more milk produced per cow. The intention of this report
is to identify what efficiency opportunities look promising for the future and to give some
consideration of what is required for the Australian dairy industry to capture these
opportunities.
There are ongoing and incremental improvements in plant and animal breeding that
influence productivity growth at an individual farm level. However, this is often limited by
factors not controlled by the farm business manager. For example, seasonal conditions
influence pasture production and therefore milk solids output as well as the extent of
supplementary feeding that reduces the efficiency of the relative input to output
relationship. In addition, as farmers look to maximise profit in any given year by assessing
the relative price of inputs and milk output, a temporary distortion of productivity gain can
occur. This is especially so where prices are high and additional inputs may be used to deliver
the most profit.
15
To ensure this report focuses on identifying real improvement, productivity has been
considered as improvement in the ratio of inputs used to produce a quantity of output.
Growth in productivity will therefore occur if either output is maintained while inputs are
reduced, or output is increased with the same level of input. The main input and outputs
that affect the dairy farming business are considered to be labour, farmed hectares,
purchased supplementary feed and cows and milk solids produced. Focusing on these
variables allows the effect on productivity to be understood and quantified.
Historical growth in productivity can be measured and presented as total factor productivity
growth. Research by the Australian Bureau of Agricultural and Resource Economics and
Sciences calculates that dairy total factor productivity growth averaged 1.6 per cent per year
between 1978-‐79 and 2011-‐12 (Gray, 2014). A Victorian Department of Primary Industries
report states Total Factor Productivity growth of 0.1 per cent per annum from 1988-‐90 to
2008-‐09 for Victorian dairy (Karanja, 2012). These results compare poorly with broader
Australian agricultural productivity growth of 2.1 per cent per year from 1948-‐49 to 2011-‐12
(Gray, 2014). Despite large variation between individual farming businesses these results
indicate limited improvement in productivity has occurred.
The challenge to create real efficiency improvement at farm level must address the building
blocks of dairy farm productivity. These have been assessed in this report as themes around:
• Enabling greater labour efficiency.
• Boosting the quantity and quality of feed grown on a certain area of land.
• Increase milk production relative to what the cow consumes.
• Harness the power of farm data.
16
It is worth noting that there are two methods by which productivity growth could occur
(Karanja, 2012):
• Technological Change -‐ the adoption of new technology resulting in a more efficient
dairy farm.
• Efficiency Change – better business management by adoption of existing best
management practice.
Although not exclusively, this Nuffield report focuses on the potential for technological
change. The importance of productivity is highlighted by this quote; “Profitability is
determined by two factors: productivity and the terms of trade, derived as a ratio of prices
received to prices paid. As agricultural output and input prices are determined largely on
global markets, farm managers have a negligible influence over their terms of trade.
Therefore, it is only productivity that farm managers can improve through innovation in
technologies and management systems (Nossal and Sheng 2010). Hence, while profitability is
typically the objective of farm managers, they most commonly influence their profits through
changes in productivity” (Karanja, 2012).
17
Chapter 2: Improving labour productivity
One of the largest costs in the dairy farm budget is labour, both as the direct cash cost and
the input cost of an owner operator. Therefore, any assessment of overall productivity must
address the challenge of improving labour efficiency. On a dairy farm the obvious
measurement of this efficiency is the number of cows per person employed. It is in this
context that initiatives to improve labour productivity have been identified.
Increased labour efficiency has been supported by better milking shed design, larger farm
machinery and more employed labour to support the family based owner operator. The
mechanization of feeding and fodder conservation tasks with the more recent automation of
milking practices, has facilitated the trend towards larger farms with higher stocking rates.
Examples of this include a rotary dairy milking more than 300 cows per hour with one to two
people and a combination round baler with wrapper allowing one person to operate the
baling and wrapping activity simultaneously.
In the future, while the author expects the trend of increasing farm business size will
continue, the automation of processes driven primarily by labour saving motivations will
increasingly be supported by real time information that enables precise decisions.
Empowering labour with both automation and decision support not only has the potential to
increase productivity but will also:
• Attract labour with a different range of skill sets and interests.
• Improve lifestyle.
• Ensure consistent quality outcomes.
18
Automated milking systems
Most of the automated milking systems currently on the market originate in Europe where
dairy farms predominately manage the feeding and milking of cows in a housed system.
These robotic systems milk cows individually in a ‘box’ and have been adapted to work in our
pastoral conditions using a voluntary cow traffic system. This is required to achieve the best
utilization of the milking box asset.
Lely A4 milking robot
Lely offers a robotic milking solution where the cow chooses to enter a cubicle and a robot
then completes the milking and dispenses concentrate if required. Within a grazing system
the farmer will usually offer three pasture feed allocations per 24 hours that the cow
accesses by way of a drafting gate on leaving the milking robot.
The number of cows the robot can milk depends on factors such as milking time, milk yield,
voluntary cow flow, and farm layout. Lely estimates that the A4 robot system under
optimum conditions could harvest 2.5 million kg milk per person per year. Their expectation
is that the use of a robotic milking system with voluntary cow traffic would reduce labour
requirements by half in ideal circumstances (Mourik, 2014).
The key to achieving these optimum conditions and high robot utilization is good voluntary
cow traffic. Lely’s experience leads them to believe good voluntary cow flow is driven by
(Mourik, 2014):
• The feeding of concentrate based supplements at the robot.
• The strategic location of cow drinking water supply.
• The quality of grazed pasture affecting cow appetite.
19
Lely also introduced the author to the concept of both physical and mental relief for the staff
when using a robotic milking system. This view was due to the disruptive change to the
farming system when using voluntary cow traffic. The benefits of this go beyond just the
automation of physical cup attachment. The aspect of mental relief was described as
(Mourik, 2014):
• A reduced skill level for day to day farm activities being required; using a robotic
system can compensate for lower skilled and cheaper labour.
• The requirement for highly skilled operators can be concentrated at farm manager
level.
• Farm management would require less staff management time.
De Laval voluntary milking system (VMS)
De Laval offers a box milking robot system called the ‘VMS’. The capacity of this system is
quoted in optimum conditions as being able to harvest 3,000 kg milk per day (Sahlstrom,
2014). There has been a continual improvement in this milk harvest; for comparison the best
result in the year 2000 was 1,800 kg milk per day (Sahlstrom, 2014).
In this report the application of robotic milking systems is viewed in relation to labour
efficiency, however it is recognized that this is not the only consideration in a farmers
purchase decision. De Laval quotes the following list as the trends driving robotic milking
uptake (Sahlstrom, 2014):
• Labour constraints, human resources complexity and cost, especially on family farms.
• Demand for a flexible lifestyle.
• Requirement for high utilisation on invested capital.
• A trend towards business sustainability where robots can improve cow longevity,
deliver consistent food safety standards and optimise output.
• Farmers wanting more time to manage the herd.
20
Despite these box milking robotic systems having been available in Australia for at least 15
years, widespread adoption has been limited. Currently robotic milking installations number
approximately 28 farms across Australia with another six in construction (Kerrisk, 2014). The
author speculates this limited adoption is the result of:
• The relatively high capital investment compared with traditional dairy design,
especially where multiple boxes are required to milk more than 300 cows.
• The limited capacity of the box units requires voluntary cow traffic, necessitating
whole farm change.
• Despite more farmer flexibility, without fixed milking times the extended operational
hours of the system can result in the farmer feeling unable to ‘shut off’.
• A reluctance of many farmers to rely on a technology they do not understand and
they often question their ability to master the complexities of operation.
It is for these reasons the author expects that while new box installations will continue
where suited to individual preference, wide scale adoption over time is unlikely to occur.
Robotic milking solutions that facilitate high labour efficiency, whilst at the same time suiting
grazing systems and being cost competitive for larger herds, are described below.
De Laval automatic milking rotary (AMR)
The De Laval ‘AMR’ is a relatively new concept where multiple robotic arms attach and
remove milking cups in an internal rotary dairy design. The AMR is built with 24 bales and
has a milking capacity of 70 -‐ 90 cows per hour (Sahlstrom, 2014). When using voluntary or
semi-‐voluntary cow traffic one AMR system can milk 700 – 800 cows on a daily basis. De
Laval estimates that it would require 11 VMS boxes to match the AMR’s capacity. Currently
there are several AMR’s in Australia with another about to be commissioned.
21
Advantages:
• One installation allows the farmer to increase herd size without adding additional
boxes.
• For large herds it is a more economic proposition than multiple box units.
• The operator is not required at the AMR for milking.
Disadvantages:
• The rotary with stationary robot arm arrangement results in each cow having a one-‐
off cup attachment opportunity.
• Requires the whole farm change to voluntary or semi voluntary cow traffic to milk
more than 250 cows.
• The internal rotary design requires an additional installation in order to feed
concentrate.
Photo 1: De Laval Automatic Milking Rotary at Camden, NSW, Australia.
Source: A Pellett August 2014.
22
GEA DairyProQ
The DairyProQ robot system from GEA Westphalia has a robotic arm attached to each bale
on an external rotary platform. As the cow enters the bale and travels with the rotating
platform, the robotic arm will in a single attachment clean the teat, pre-‐milk stimulate the
teat, perform milking and then post-‐milk teat spray before the arm releases.
In designing this world first system GEA identified the challenges they wanted to overcome
(Hille, 2014). These included:
• Finding and retaining labour on large farms.
• Ensuring food safety.
• Achieving high standards for animal health.
• Developing technology that can operate reliably 24 hours a day.
It was after this analysis that GEA started to develop the DairyProQ robotic system in 2012.
Equipping each bale with its own robotic arm allows fewer incomplete milkings compared
with a fixed robotic arm arrangement and the rotary size can be built to achieve the desired
throughput. Importantly, this system ensures in-‐built reliability because any robotic failures
can be isolated on one bale and do not stop the whole system operating. A rotary platform
using the DairyProQ system requires the farmer to supervise the milking process.
The hourly milking capacity of a DairyProQ system varies from a 28 bale rotary milking 130
cows per hour to an 80 bale rotary milking 400 cows per hour (Hille, 2014). The application
of this technology to the Australian environment offers significant benefits. This is the first
robotic system that is capable of batch milking. The ability to batch milk using a robotic
system is highly desirable for Australian conditions because:
• The farmer retains active control over pasture and forage management.
23
• It captures the physical stress relief and information capture of a robotic milking
system, without the whole farm change required with voluntary cow traffic.
• The capacity to feed concentrate during the milking process.
• The ability to actively manage the milking frequency of late lactation cows and in
adverse weather conditions.
This system has the potential to substantially increase the number of cows one person can
milk, while retaining the farm system that already delivers optimum pasture harvest and the
peace of mind of a system not operating 24 hours per day. As this report is written, the
DairyProQ system is expected to be available in Australia within two years. Whilst pricing is
commercially sensitive it is likely to compare favourably to multiple box systems setup to
milk 400 cows or more (Burgt, 2015).
Photo 2: First commercial DairyProQ installation in Germany.
Source: A Pellett September 2014.
24
Robotics
The use of robotics in a dairy farm context has until now been limited to the milking process
in grazing systems. Other robotic applications suitable for dairy farm activities are becoming
feasible as technology develops and cost declines. The basic robotic hardware is often
developed in other industries and requires adaption for grazing dairy systems.
Fetching robot
The Australian FutureDairy research and development program has created a customized
robot capable of herding cows from the paddock autonomously. This is achieved using a
generic skid steer platform equipped with a dairy cow specific algorithm.
Photo 3: Autonomous Robot trialled by FutureDairy in Camden, NSW, Australia.
Source: A Pellett August 2014.
25
The basic machine is a Clearpath Robotics™ product called a Grizzly (Clearpath Robotics). It is
able to navigate the challenging dairy farm environment with electric four wheel drive, a 3 –
12 hour run time, 200mm ground clearance, 80 horse power and a robust frame that
enables up to 600 kg of additional equipment to be mounted on the robot. The FutureDairy
team has equipped and programmed this multi-‐purpose industrial robot to ‘see’ cows in a
paddock and muster them as a herd (FutureDairy Project).
Using this type of platform there is the potential to develop additional functionality that
automates other farm tasks. The FutureDairy project manager suggests it would be possible
to extend the robot’s capabilities to include (Kerrisk, 2014):
• Around the clock monitoring and reporting on calving events.
• Pasture quality and quantity assessment.
• Conduct routine paddock checks such as water supply, electric fence power.
• Soil sampling with GPS reference.
• The spot spraying of paddock weeds.
• Targeted fertiliser application related to soil test information and pasture
assessment.
There are obvious savings in labour where farm activities can be automated. The use of
robots with smart dairy-‐specific algorithms extends this benefit to include:
• Improved animal welfare, as a robot is not time constrained.
• Consistency and quality of farm activity.
• The capture of paddock information can be used to adjust fertiliser application based
on yield and soil fertility, as well as more accurate pasture allocation.
• The analysis of animal observation information could facilitate better culling
decisions, reduce animal health costs, and enable earlier treatment leading to
improved animal performance.
26
Heavy activity paddock robots
The seasonal nature of pasture production relative to a farmers stocking rate necessitates
time consuming fodder conservation and feeding activities. Robotic and autonomous
tractors with implements have the potential to boost labour productivity.
An example of this capability is the autonomous mower currently in development by the
Autonomous Tractor Corporation. This concept is based on a modular engine and drive unit
customized for the work required (Tobe, 2014). Using robotic tractors for paddock activities
also allows size and power requirements to reduce due to the continuous working potential
of a robot. In addition to labour savings there is also lower fuel consumption and less risk of
larger machinery causing ground compaction.
Conclusions
The introduction of the robotic rotary will significantly boost labour productivity of the
milking process by allowing one supervisor to milk more than 500 cows, providing physical
stress relief and individual cow information. The ability to batch milk cows provides a
solution more suitable to the Australian context at a price comparable to the installation of
multiple box systems.
Development of dairy-‐specific algorithms that control generic paddock robots will further
expand the supervisory role of one operator. These are most likely to be developed within
the Australian industry, due to the unique combination of large scale grazing systems and
high labour costs, providing a catalyst for change.
The benefits of robotic technology extend beyond the automation of farm tasks. Robotic
systems capture information that can be used to reduce costs, ensure quality, improve
animal welfare, improve lifestyle, and increase yields.
27
Chapter 3: Higher yields from home grown feed
One of the most important factors contributing to a dairy farmer’s profit is the amount of
home grown feed consumed per hectare and per cow. In the Irish environment the
relationship between increasing net profit and improved pasture consumption per hectare is
calculated at 43 per cent (O'Donavan, 2014). The predominant feed source grown on farms
in Victoria and Tasmania is ryegrass. As the quantity and quality of home grown forage
increases, productivity improves as the farmer chooses the desired combination of increased
stocking rate, more milk per cow or reduced purchased feed.
Enhancing the productivity of ryegrass
The DairyFutures Cooperative Research Centre (CRC) continues to develop a range of
technologies that will, as Dr. David Nation describes, “address a fundamental business
imperative for dairy farmers: making a profit from the biology of growing grass and
producing milk”. The CRC’s designer forage program aims to deliver a more productive,
nutritious and resilient forage base (Dairy Futures CRC, 2014).
Genomics assisted breeding
Knowledge of ryegrass Deoxyribonucleic acid (DNA) markers provide plant breeders with
information on the diversity that is contained within existing elite cultivars. For instance, the
yield range for individual plants can be very large and range from 50 to 250 grams (Nation,
2014). This illustrates the potential for targeted plant breeding using DNA markers for traits
such as digestibility and yield.
The application of genomic selection by commercial plant breeding companies has the
potential to double the rate of genetic gain in yield and persistence. This is largely because
the typical breeding timeline of twelve years can be shortened to six years. The use of
genomic information reduces the cost to bring a new ryegrass to the market as computer
simulations reduce the number of field trials required (Dairy Futures CRC, 2014). As the
28
development of new cultivars is driven by commercial consideration, this technology with its
reduced cost may allow profitable development of varieties suitable for specific Australian
conditions.
Hybrid ryegrass
Unlike other crops (such as maize) creating a hybrid ryegrass plant has not been possible due
to ryegrass self-‐fertilisation of seed production. The CRC is working on technology to
overcome this barrier. Genomic selection of two unrelated seed lines could then be crossed
(with 83 per cent seed set) producing a ryegrass variety benefiting from yield increases of 20
per cent (Nation, 2015). This hybrid vigour would remain for the life of the plant as seed set
in commercial dairy pastures is not desirable.
Transgenic high energy ryegrass
The CRC has developed a genetically modified line of ryegrass developed with improved
digestibility rates. The advantage of this ryegrass is that it gives similar or better yields to
commercial cultivars but with available energy being one megajoule/kg higher. Significantly,
the higher digestibility occurs in seasons outside of spring, when digestibility is most limiting
to animal performance (Nation, 2015). For farmers the significance of this technology is to
give a potential 10 per cent increase in milk solids produced per hectare, or reduced
purchased supplementary feed.
It is beyond the scope of this report to debate the merits of genetic modification (GM) within
the food chain. However, it is worth noting the significant challenges facing the
commercialization of this ryegrass line:
• Will customers buy milk products where a large percentage of a cows diet has been
declared to include genetically modified plants, and if so, at what price?
29
• Despite a scientific message defending GM, will society consent to farmers using this
technology where the perceived benefit accrue exclusively to the farmer?
• How to make a business case for the Australian market to commercialize the GM
ryegrass given the major seed companies are New Zealand based with a blanket ban
on GM.
Endophyte selection
Endophytes are fungi that co-‐exist with grass plants living between the cell walls. They
produce alkaloids that can provide the grass plant with protection from insect pest damage.
Most grasses that are not selected with an improved endophyte have a naturally occurring
wild endophyte. These wild endophytes contain a range of toxic alkaloids such as Lolitrem b
that can cause grass staggers, and Ergovaline, which can cause heat stress.
In New Zealand, Cropmark Seeds has been breeding hybrid grasses called festuloliums for
many years. One of the benefits of these ryegrass meadow fescue hybrids is that meadow
fescue can contain an endophyte called Neotyphodium uncinatum which is not found in
perennial ryegrass. This endophyte has no Lolitrem b or Ergovaline but has alkaloids called
Lolines. Available through Cropmark Seeds the commercially available ‘GrubOUT U2™’
endophyte (a strain of Neotyphodium uncinatum) has several advantages. This Loline
containing strain is safe for animals, toxic to a wide range of pasture pests, and is found in
both the roots and leaves, unlike other endophyte (Cameron, 2014).
Pasture inoculated with the Loline endophyte is afforded protection from a range of pests
including (Cropmark Seeds, 2014):
• Grass grub larvae.
• Black beetle adults and larvae.
• Argentine stem weevil (adults and larvae).
30
• Porina caterpillar (lolines kill porina caterpillar in forced diet situations).
• Red headed pasture cockchafer.
• Black field cricket.
• Root aphid.
• Pasture mealy bug.
• Soil nematodes.
Photo 4: Cultivar Field Trials, Ashburton, New Zealand.
Source: A Pellett December 2014.
It is worth noting that the CRC is also developing new endophytes. Research has focused on
identifying alkaloids that transmit successfully to different ryegrass cultivars and introducing
a specific mutation designed to delete the toxins responsible for negative animal
performance (Nation, 2015). The comprehensive knowledge about endophytes now makes it
possible to design an endophyte for specific pastures and pests. The alternative approach is
31
far more time-‐consuming and expensive. For comparison, Cropmark Seeds last winter tested
85,000 individual plants for endophyte proteins.
The significance of selecting grass cultivars with new endophyte in enhancing productivity is
threefold:
• Pasture yield is not constrained by pest and insect attack.
• Animal performance is not constrained as it might be with wild endophyte.
• The likely improved pasture persistence reduces the forced requirement for pasture
renovation, reducing cost and allowing poorer performing paddocks to be targeted.
Cultivar selection tools
Recently both Ireland and New Zealand have launched comparison indexes to help farmers
choose the best performing grass cultivars. Cultivars are ranked based on a calculated
economic value and presented in an index format. This type of system is planned for
Australia and is in development by Dairy Australia (Romano, 2015).
The pasture profit index -‐ Ireland
The Irish dairy industry has a goal of increasing 2007 – 2008 milk production 50 per cent by
2020. One of the strategies to achieve this is by lifting pasture utilisation from an average 7.3
tonnes per hectare to 10 tonnes per hectare by 2020 (O'Donovan, 2014).
The recent introduction of the Pasture Profit Index by the Irish Agriculture and Food
Development Authority is an initiative designed to deliver on the strategy. It is a total merit
index designed to assist grass cultivar selection by farmers. Important grass performance
traits such as seasonal dry matter production, quality, persistency, and silage yield are given
varying economic values. Cultivars on a recommended list are given a ranking based on total
economic merit (O'Donovan, 2014).
32
Table 1: Calculation weightings in the Irish Pasture Profit Index
Factor Dry Matter Yield Quality Silage Cuts (1st &
2nd)
Persistence
Weighting 40% 15% 10% 35%
Pasture profit index -‐ Euros per ha per year
NB: Seasonal Dry Matter Yields are valued Spring 0.16, Summer 0.04, Autumn 0.11 in Euros
Source: (M. O'Donovan, 2014)
The information provided in this index is supported by another Irish initiative called the
National Grassland Database and branded as PastureBase Ireland. This database is used by
farmers to record their pasture assessments taken at 16 day intervals and with cultivars
identified per paddock. With some 400 registered dairy farmers, half are regular users. The
information is used to help benchmark the pasture management of the industry and provide
relative cultivar performance using commercial farm data (O'Donovan, 2014).
The DairyNZ forage value index -‐ New Zealand
Since 2012 New Zealand dairy farmers have been able to use the Forage Value Index to help
support ryegrass cultivar selection. The index is produced by DairyNZ in conjunction with the
New Zealand Plant Breeding and Research Association. Farmers can access the index via a
web-‐based tool with the ability to filter selections on endophyte, geographical location,
ploidy, ryegrass term and heading date. The addition of quality and persistence information
is planned when evaluations can support this information (Bryant, 2014).
This index calculates an economic value presented in five star rankings, representing dollars
per hectare. The economic value is calculated using a model called Farmax Dairy Pro and is
33
revised every year. This incorporates the effect of increased dry matter weighted by season
and region on; animal performance, conserved feed and supplement saved. The index also
provides seasonal performance values for winter, early spring, late spring, summer and
autumn and are given different values depending on the region of New Zealand (DairyNZ,
nd).
Table 2: New Zealand Forage Value Index Report Example.
Source: DairyNZ website, feed section, cultivar selection, 2015.
$388 to $514
$262 to $387
$135 to $261
$9 to $134
-$117 to $8
Filtered by: UPPER SOUTH ISLAND, PERENNIAL RYEGRASS/AR1/AR37/NEA2/ENDO5/LE/SE/WE, DIPLOID,VERY LATE
FVI STAR RATING ($/HA)
Using the Forage Value Index (FVI), this tool allows you to objectively select the mostsuitable ryegrass cultivar for your farm. Select the relevant filters to find the cultivarsthat best fit your criteria. Teal = selected. Grey = unselected
Forage Value Lists for download:Click on the region links below to open and download the forage value list.
Perennial Ryegrass: Upper North Island , Lower North Island , Upper South Island , Lower South Island
12 Month: Upper North Island , Lower North Island , Upper South Island , Lower South Island
Winter Feed: Upper North Island , Lower North Island , Upper South Island , Lower South Island
MATRIXSE 9.7 3 4 2 4 4 SE Diploid
VeryLate
Cropmark
PERFORMANCE VALUES
CULTIVAR FVI (STAR
RATING)
CONFIDENCE WINTER EARLY
SPRING
LATE
SPRING SUMMER AUTUMN ENDOPHYTE PLOIDY HEADING
DATEMARKETER
!! !!
! !! !!
! !! !
! !!
! !
!
CULTIVAR SELECTOR TOOL - FVI
34
Fodder beet
The attractiveness of growing fodder beet is its high yield potential. When grown
successfully, yields of 20 -‐ 30 tonnes of dry matter (DM) per hectare can be achieved
(Pedofsky, 2014). For this reason it has grown in popularity in New Zealand since 2007 where
it has been used to supplement lactating cows and as a winter feed source for dry cows and
young stock.
Along with high yields the characteristics of fodder beet include (DairyNZ, nd):
• A high energy feed at approximately 12 MJ / kg DM.
• Low protein levels of 13 per cent.
• Low fibre levels of less than 20 per cent.
• High sugar content making the crop very palatable.
In New Zealand the potential for fodder beet is being assessed by the Beef and Lamb funded
‘Fodder Beet Profit Partnership’. This is a group of 37 sheep, beef and dairy support farmers
collectively sharing their fodder beet experiences. The program establishes yields, gross
margins, growing cost, animal performance, and potential pitfalls. The information
generated is used by farmers and industry providers to inform cultivar selection, refine
agronomic practices, ensure maximum animal performance, and establish the crop’s best fit
within the production system.
The benchmark first year figures for 2014 indicate (Beef and Lamb New Zealand, 2014):
• Average yield of 18,200 kg DM per hectare (range 11,657 -‐ 27,800).
• Average cost to grow of NZ $2,480 per hectare (range $1,865 -‐ $3,451).
• An average cost of production of 14 cents per kg DM grown (range 8 – 26 cents).
35
NB: These results are for Canterbury foothill farms without irrigation at an average altitude
of 305m, 37 different sites are included.
Photo 5: Farmers Fodder Beet Sowing Date Trial, Windwhistle, New Zealand.
Source: A. Pellett December 2014
Conclusions
Ryegrass is likely to remain the base of the Australian dairy grazing system. The ability to
speed up improved cultivar development at reduced cost is an opportunity to place more
emphasis on selecting grasses that perform best in Australian conditions rather than the
broader Australia and New Zealand marketplace.
36
The inoculation of the best performing cultivars with novel endophytes will boost pasture
production and animal performance. Pastures that persist for longer will allow a targeted
selection of under performing pastures to be renovated lifting overall farm feed production.
Tactical use of fodder beet can increase overall farm feed production at a comparatively low
cost of production.
As new breeding technologies enhance ryegrass performance, the widespread use of a
pasture index in Australia will result in the best performing pastures for different needs
being demanded. Plant breeders would be encouraged to focus research efforts on even
more profitable cultivars as the index drives cultivars commercial success.
37
Chapter 4: Precision dairy
In the future it is clear that technology will become more important for the dairy farm
business. The use of a broad range of technologies for monitoring and management can be
described as ‘Precision Dairy’ or ‘Smart Dairy Farming’. The productivity benefits of this
emerging theme are a combination of:
• Increasing yield of both pasture and animals.
• Reducing costs as inputs are more targeted and timely.
• More efficient labour with reduced manual observation.
The term precision dairy has been defined as: “The use of information and communication
technologies for improved control of fine-‐scale animal and physical resource variability to
optimise economic, social and environmental dairy farm performance” (Eastwood, 2008).
The developing functionality of precision dairy goes beyond what, until recently, has largely
been the automation of farm tasks focused on labour savings within the dairy shed. Now and
in the future, intelligent use of information collected from individual animals, pastures, and
soils will allow farmers to optimise production with far greater efficiency than has been
achieved. Developments in this area will help address the challenges created by growing
farm size, scarcity of labour, and complex production systems.
While the technology enables the collection and analysis of information, productivity gain
will only occur when the farmer makes timely decisions.
38
Cow sensors
Photo 6: University of Kentucky Research Dairy Farm, Sensor Field Testing, United States.
Source: A. Pellett, December 2014
A range of available animal sensors capture information in the categories of (Bewley, nd):
• Nutrition -‐ variable in dairy feeding based on health, size, and yield.
• Production.
• Health -‐ mastitis, rumen condition, metabolic disorders, body temperature.
• Fertility – heat detection, pregnancy, and calving notification.
As an example of the range of options available, a comprehensive but non-‐exhaustive list of
technologies can be found by accessing the web address:
http://afsdairy.ca.uky.edu/files/extension/decision_aids/precision_dairy_technology_list.xls
m
39
DairyMaster MooMonitor+
This collar based cow monitor developed by DairyMaster in Ireland captures both cow health
and fertility information with an algorithm designed to work in grazing systems. The
MooMonitor+ is able to record when the cow is in heat, her feeding time, resting time, and
rumination activity every 15 minutes. This information is transmitted up to one kilometre
throughout the day, or saved for later upload as cows enter the dairy where a reader is
present. The intensity of this measurement activity results in 210,000 data points per cow
per year. The farmer can access the information either through a computer or mobile phone
application, providing the ability for real time notifications. The MooMonitor+ software can
draft cows automatically and has the added functionality of audible voice alerts during
milking (Harty, 2014).
Using this type of information, in conjunction with milk production parameters, can add
validity to a system-‐generated prediction. For example, confidence in a predicted mastitis
event would be higher where cow activity, feeding / resting time, and milk production
parameters are combined compared with using milk conductivity alone. The consolidation of
data from multiple sensors will considerably improve decision support accuracy compared
with single source technology.
Some insightful quotes highlight the relevance of precision dairy:
“There will be more focus on the individual cow in the future” (Harty, 2014), and
“The future of disease detection, should you check the cow or check the computer?”
(Bewley,nd).
Herd navigator
The herd navigator product was launched in 2008 and is marketed by De Laval. It is yet to be
released in the Australian market but is sold in many northern hemisphere markets. This
40
product can provide automated on-‐farm analysis of milk. This is combined with De Laval
milking technology to form an automatic diagnostic tool and herd management solution.
The system is programmed to take samples from cows as they are milked automatically.
These samples are selected only for cows that are due to be tested, from particular milking
events, and on particular parameters. These samples can be tested on farm for (Herd
Navigator):
• Timing of heat, pregnancy, and abortion by measuring progesterone.
• Mastitis, both clinical and subclinical, by testing the prevalence of an enzyme called
LDH.
• Energy balance by testing for ketone levels.
• Diet optimisation by measuring urea indicating energy and protein balance.
This type of system has the potential to automate, standardise, and improve the accuracy of
heat detection, pregnancy testing, mastitis detection, and identification of poor appetite and
cow health events. Farmers benefit by treating cows earlier, thereby incurring less lost
production and reducing the incidence of costly clinical cases. Automatic sampling of milk
throughout a cow’s lactation not only makes heat detection easier but removes the need for
invasive procedures commonly used to confirm pregnancy.
Precision dairy in the paddock
The ability to remotely sense pasture yield and quality would be extremely useful. Receiving
this type of information throughout the growing season would enable more accurate pasture
allocation decisions. A more complete knowledge of what feed is available, at what quality
and how fast it is growing would facilitate better grazing management.
41
The combination of Remote Piloted Aerial Systems, GPS navigation, and sensors has the
potential to deliver pasture yield and quality measurements (Yule, 2014). Research at
Massey University in New Zealand is being conducted using high definition imagery taken at
660 metres above ground in 350 metre widths. At present, quantifying pasture covers
requires a lot of data processing. For example, in order to get high resolution, a pixel size of
2.5 centimetres on the ground equates to 16 million pixels per hectare requiring analysis. As
technology continues to improve this approach has the potential to automate pasture
assessment. As well as replacing the often time consuming manual methods, using high
definition imagery may improve accuracy and consistency. Using this information in
conjunction with ground based robots would enable automated weed control and variable
fertiliser application.
Virtual fencing is another technology that has potential within the precision dairy area
(Trotter, 2014). Control of pasture allocation can be achieved using a special collar worn by
the herd. The farmer can allocate pasture by setting a virtual fence that the animal respects
when warned by audible alarm and electrical charge. The application of virtual fencing could
make multiple daily allocations possible and, when linked to remote pasture assessment,
help manage grazing. In addition, for farmers with voluntary cow grazing systems, these
collars may assist in prompting cow traffic flow.
Precision dairy challenges
With the adoption of most new technologies comes more data that can be turned into
management information. However this additional information is only beneficial if farmers
make better decisions. The most attractive technologies to farmers are labour saving in
nature as these demonstrate a tangible cost/benefit comparison. Benefits gained from
adopting technologies that increase efficiency can be more difficult to calculate and require
a longer time frame to recover the investment (Edwards, 2014).
42
As technology rapidly evolves, any investment in new technology is quickly out of date.
There is reluctance to invest in any particular system due to the risk of the purchased
technology becoming quickly outdated. This could necessitate a new investment model such
as a subscription based financing model. These short term leases on new machines and user
information systems would ensure continued access to the latest technology (Smit, 2014).
Capturing, analysing and controlling data is becoming increasingly important. The
information used to make optimum decisions is often generated within systems provided by
private companies. This results in a number of challenges:
• Data from different systems cannot be aggregated and lacks standardised data
definition.
• Knowledge is isolated from industry good organisations and accrues to private
companies.
• Switching systems is difficult due to the level of investment and customised
installation.
Optimising the value chain
The abundance of information generated by precision dairy technologies on farm in real
time has potential to support a more responsive value chain.
Demand for farm inputs such as fertiliser could be predicted based on paddock yield, soil
fertility mapping, historical application and forecast weather events. As a consequence of
this prediction the timing of supply and price may adjust accordingly. In the Netherlands a
program has been developed where the robotic milking system, having detected a cow is
ready for breeding, provides the artificial breeding company with historical genetic
information and production data that determines sire selection and requests an artificial
43
insemination (AI) technician (Roos, 2014). The farmer contracts a range of bulls at a fixed
price and the service offering becomes integrated into the farm business.
Improved accuracy of forecast milk production would be of great value to the milk
processor. Precision dairy technologies will allow more accurate predictions of future supply
patterns based on the real time analysis. The ability to combine data such as milking cow
numbers, lactation stage, production, pasture growth rates, soil temperature, soil moisture,
weather forecast and supplement prices could be part of the milk supply arrangement
between farmer and processor. Milk processors armed with this information could optimize
product mix in response to customer demand. Development of dynamic price signals
between the processor and farmer would encourage farm production to be adjusted
according to marginal profitability much faster. The author expects that the faster matching
of supply and demand through the supply chain would reduce price shocks.
Another benefit is that customers can more easily make purchasing decisions based on
information collected on-‐farm. The process of guaranteeing quality is automated and the
information generated provides traceability through the supply chain to the farm.
Conclusions
The development of precision dairy technology is in its infancy; it is likely both the
cost/benefit and accuracy of such systems will improve over time.
Harnessing data from multiple sources on the farm improves the accuracy and confidence in
precision dairy decisions. These systems can increase yields, reduce costly inputs and allow
labour to focus on higher value management activities.
44
The improved capacity to generate farm based information in real time offers the potential
to change the way the supply chain is integrated and how fast the demand and supply
equation is balanced.
If Australian dairy is to enjoy the full benefits of precision dairy, investments need to be
made to ensure the nature of the grazing system is considered. It would be sensible to
collaborate with other countries focused on grazing systems.
45
Chapter 5: Cow efficiency
For most dairy farmers livestock represent the second largest business asset, therefore any
increase in animal efficiency will provide significant productivity gain. With purchased feed
being a major cost item any improvement in dairy cattle feed conversion efficiency would be
of major benefit. This would allow feed costs to be reduced while holding production
constant; more likely in a grazing system, an increase in stocking rate translating into more
milk solids per hectare.
While animal efficiency is a major focus in the pig, poultry and beef industries and despite
large increases in dairy cow total production, selecting for feed conversion efficiency has
proved difficult in lactating grazing cows. Accurate assessments have proved difficult due to
variation in rumen fill, changing bodyweight, and body reserves that are mobilized during
lactation. The lifetime efficiency of a lactating dairy cow is also a more complex breeding
goal than that of growing a pig, chicken or beef animal to a target weight (Macdonald, 2013).
What is cow efficiency?
Gross efficiency
The widespread use of artificial insemination to access improved genetics combined with
better nutritional management of cows has contributed to increased milk production
relative to bodyweight. For example, in the New Zealand dairy industry, the Livestock
Improvement Corporation (LIC) estimates a one per cent gain per annum. This is a measure
of gross efficiency achieved through dilution of the fixed costs of maintenance, while
increased milk production drives increased feed intakes. The potential for ongoing
improvement in gross efficiency is finite and is assessed to be four per cent of live weight as
a maximum intake of pasture (Davis, 2014).
46
Feed conversion efficiency or residual feed intake
Greater feed conversion efficiency (FCE) does not necessarily relate to total levels of milk
production. FCE is the difference between the actual amount of feed eaten and the expected
amount eaten, this is referred to as residual feed intake (RFI) (DairyNZ, 2013). Of interest to
the dairy farmer is that some cows, although of the same size, equivalent milk production
and achieving similar reproductive performance, require less feed than other cows.
Is the RFI difference significant?
Research on RFI has recently been completed by very similar studies conducted in parallel in
Australia and New Zealand.
The New Zealand study evaluated 80 cows for feed conversion efficiency over 35 days in
early first lactation. The cows were measured in two groups of 40, classified as most and
least efficient using the results of an earlier heifer calf study (Macdonald, 2013).
Table 3: Milk solids production and feed intake for high and low efficient cows
Most efficient Least efficient SED P
Milk solids (kg/cow/day)
1.05 1.04 0.01 0.82
Mean live weight (kg)
407 401 6 0.31
Feed intake (kg) DM/cow/day)
18.9 19.1 0.32 0.39
Divergence# (kg DM/cow)
-‐0.31 0.31 0.22 0.007
#Divergence is the difference between the high and low efficiency groups
Source: (Waghorn, 2013)
47
The significance of the range in dry matter intake required to produce the same amount of
milk is that the most efficient cows require three to four per cent less dry matter.
The study in Australia was conducted in Victoria and measured 50 cows classified as feed
efficient and 58 cows classified as feed inefficient based on the results of the heifer calf
study mentioned earlier. These animals had RFI calculated in first lactation for 32 days by
assessing dry matter intake of cubed lucerne hay and grain, while measuring milk production
and live weight change. The findings of this study confirmed the New Zealand results.
Differences in feed conversion efficiency identified as calves remained significant once the
same animal was a lactating cow.
Is it a feasible breeding objective?
The difficulty in including RFI in a breeding program has been the high cost of widespread
recording of feed intake in conjunction with changes to cow live weight and milk production.
The impracticality of measuring these factors in a widespread progeny test program means
that validation of RFI using actual sire daughter information is currently unlikely.
In New Zealand, LIC conducted an experiment to see if RFI could be predicted in a group of
independent cows. Genotyping of 3359 cows born in 2005 or 2006 was completed using the
Bovine SNP50 BeadChip. From this group 197 cows (representing the top and bottom 10
percent) were successfully evaluated for RFI after having been selected using genomic
predictions for RFI.
These results show that dairy cows selected using a genomic prediction for RFI
demonstrated actual differences in RFI. The use of genomic technology to form a prediction
of RFI has allowed the recent development of the world first ‘feed saved’ Australian breeding
48
value. This is expected to be released within a breeding index during April 2015 (Dairy
Australia, 2015).
Table 4: Actual difference in RFI when selected for RFI by genomic prediction
RFI Group Low High SED
Efficiency (efficient) (inefficient)
N 99 98
Milk solids kg/day 1.38 1.34 .03 ns
Milk yield kg/day 16.8 16 0.5 ns
LW kg 514 511 5.6 ns
LW change kg/day 0.24 0.27 0.02 ns
DM intake kg/day 25 25.6 0.27 ns
RFI kg DM/day -‐0.35 0.36 0.22 ***
(LW = live weight, ns = not significant, *** P<0.001)
Source: (Steve Davis, 2013)
Can RFI fit within a broader breeding objective?
As farmers have much broader breeding goals than just achieving a more feed efficient cow,
selection for RFI will have to sit alongside many other desirable traits within a breeding
program. Defining an industry-‐wide breeding objective is increasingly important as the rate
of genetic gain speeds up due to genomic technology. Failing to define an all encompassing
49
future-‐looking breeding objective could have negative impacts on future dairy herd
profitability (Berry, 2014).
Inclusion of RFI within a breeding objective will have some impact on the progress of other
traits being selected for. This is the case even when RFI is independent of milk production as
the selection intensity for other traits is reduced to accommodate an additional breeding
objective.
Leading geneticist Donagh Berry suggests that a national breeding goal should be holistic and
future-‐looking and nominates the following characteristics defining the ideal cow of the
future (Berry., 2014):
• Produces a large quantity of milk (and meat).
• Good reproductive performance.
• Good health status.
• Good longevity.
• Does not eat a large amount of food.
• Easy to manage.
• Good conformation.
• Low environmental footprint.
• Resilient to external perturbations such as extreme weather and new diseases.
The message is not to breed for one characteristic in isolation and include traits that are
likely to become more significant in the future.
It is important to note that although these characteristics will vary in heritability, the ability
to achieve genetic gain is less reliant on the strength of heritability, thanks to the use of
genomics. Rapid genetic gain in lowly heritable traits is possible. This is dependent on high
50
accuracy of selection and genetic variation. The extent of genetic variation across all traits is
similar at approximately five per cent (Berry, 2014).
A method to improve accuracy of selection for the RFI trait collection of data from a large
population at reasonable cost is required. A solution could be the profiling of dairy cows
using the results of milk samples. Milk mid-‐infrared spectroscopy is regularly used for testing
fat and protein and has the potential, once combined with prediction equations, to quantify
feed intake, cow energy balance, feed efficiency and methane emissions (Berry, 2014).
Broader herd productivity
Genomic prediction
In the United States farmers can use a genomics-‐based product called ‘Clarifide’ that
provides predicted animal profitability in a net merit index to inform breeding and
management decisions (Zoetis, 2012). For example, heifers with predicted profitability less
than the herd average could be culled before incurring rearing costs or sold through the
export market in the knowledge that the herds best genetics have been retained. Elite
heifers could be identified and submitted to a more costly sexed semen mating strategy
instead of a blanket approach.
Precision genomic mating -‐ designer matings
Once a broad breeding goal has been developed, the use of genomic information can assist
in facilitating rapid genetic gain by highly targeted matings. This approach effectively
predicts the additive merit of mating individuals, including which traits are likely to present
themselves as dominant in the resulting progeny. This would allow selection of a sire with
the best combining ability for individual or groups of cows with the guarantee of which traits
the next generation will inherit and display (Berry, 2014).
51
Concerns about negative animal performance caused by inbreeding can also be reduced
using genomic information in breeding decisions. This is due to the variation between the
proportions of genes shared by related animals. Although very unlikely, two full siblings can
be completely unrelated. Genomic information could be used to retain desirable elite
genetics when breeding related animals. There would be less need to restrict breeding
choices to eliminate undesirable inbreeding effects as genomic knowledge assures that
blanket eradication of lethal genes is not required (Berry, 2014).
Genomic guided farm management
The term ‘reprogenomics’ is defined by Donagh Berry as how the genome of an animal
affects its response to alternative reproductive treatments. This implies the use of genomic
information to select what would be the most effective reproductive strategy for an
individual cow. Productivity would benefit from the targeted use of synchronisation and
sexed semen, in the knowledge that application of these more expensive approaches is used
only where the cow is likely to have the best chance of pregnancy.
Understanding what animal health status a cow is predisposed towards by using genomics
provides an opportunity to tailor preventative treatments. An example is the use of dry cow
treatments only on predicted high risk cows.
As genomic predictions become more accurate, the richness of information will enable cows
to be bred selectively for differing production systems. Management decisions would also be
more informed and successful when based on the unique knowledge of a particular herd or
individual. This may improve calving rates in seasonally calved herds as the use of controlled
internal drug release (CIDR’s) are guided by the predicted responsiveness of the individual
cow.
52
Conclusions
Differences in feed conversion efficiency between lactating cows exist and are economically
significant. Using a proving scheme for the trait in commercial herds is difficult and
expensive. Currently the opportunity to breed for better feed conversion efficiency is guided
by genomic prediction.
It is recommended that selection for better feed conversion efficiency is part of a wider
breeding objective. Breeders should recognize that the use of genomic information can
significantly speed the rate of genetic gain and hence the important consideration of likely
future objectives.
Using genomics to understand an individual herd will allow the farmer to form an efficient
customized approach to herd management. Moving beyond a blanket approach will achieve
better results boosting productivity.
53
Recommendations
There is opportunity for dairy industry resources to re-‐focus research and development
activities on robotic milking systems capable of batch milking cows at commercial scale.
Greater uptake of desirable robotic technologies would be enhanced by the introduction of a
new technology access finance model. The dairy industry can play a role in proactively
developing a new model.
Assessment of the potential for fodder beet in Australian dairy conditions is required and
would be a valuable industry resource. Independent regional trials are needed to quantify
the best performing cultivars, recommended sowing dates, harvest dates, and weed
management strategies. Guidelines such as these would provide farmers with the technical
information to experiment on farm and analyse the potential fit within different production
systems.
As the range of systems available to capture and interrogate farm data expands there is a
growing challenge in consolidating data from multiple proprietary systems. Consideration
should be given to forming a dairy data co-‐op owned by farmers. This entity would enable
management of farm level information from multiple sources, provide control and security
for the farmer, be accessible for industry good research, and mitigate the cost and
technology capture by global providers.
Dairy Australia has a perennial ryegrass cultivar selection index in development expected to
be available in the next two to three years. Extending this excellent concept to an online
knowledge bank would assist farmers grow and consume more pasture per hectare.
Additional information uploaded by both participating farmer and research sites could
54
include cultivar growth rates, leaf emergence rates, endophyte type, soil fertility, soil
moisture, plant population, and pasture establishment and renovation techniques. Pasture
treatments such as fertilisers, pesticides, and gibberellic acid could easily be noted and
evaluated by farmers.
To explore and promote the possibilities of precision dairy technology the dairy industry
should establish a demonstration farm. This would be a unique model that combines robotic
and autonomous systems both in the milking process and around the farm. The focus of the
farm is to quantify the commercial benefits and determine the possibilities for the whole
farming system. Of additional interest would be the impact this farming system may have on
the wider value chain.
55
References
Beef and Lamb New Zealand. (2014). Fodder beet -‐ Our Reality. Field Day Proceedings, The
Fodder Beet Profit Partnership Canterbury, Windwhistle.
Berry, D. (2014, September 22). (A. Pellett, Interviewer) Moorepark, Ireland.
Berry, D. (2014). Breeding the dairy cow of the future -‐ what do we need? Moorepark, Co.
Cork, Ireland.
Bewley, M. (n.d.). Currently Available Precision Dairy Farming Technologies. Retrieved
February 26, 2015 from
http://afsdairy.ca.uky.edu/files/extension/decision_aids/precision_dairy_farming_technolog
ies_factsheet.pdf
Bryant, J. (2014, November 11). (A. Pellett, Interviewer) Newstead, New Zealand.
Burgt, J. (2015, February 12). (A. Pellett, Interviewer) Warragul, Victoria, Australia.
Cameron, N. (2014, December 1). (A. Pellett, Interviewer) Methven, New Zealand.
Clearpath Robotics. (n.d.). Retrieved February 23, 2015 from clearpathrobotics web site:
www.clearpathrobotics.com/grizzly
Cropmark Seeds. (2014). General Plant Breeding Research Activity. Presentation.
Dairy Australia. (n.d.). Retrieved February 11, 2015 from Dairy Australia:
www.dairyaustralia.com.au/markets-‐and-‐statistics/farm-‐facts
Dairy Australia. (2015, February 20). Same milk. less feed: World leading feed saved ABV
launched at Ellinbank. Ellinbank, Victoria, Australia.
Dairy Futures CRC. (2014). Program 1 Designer Forages. Annual Report.
DairyNZ. (n.d.). FeedRight Information Sheet. Retrieved February 24, 2015 from DairyNZ:
www.dairynz.co.nz/feed/crops/fodder-‐beet/feedright information sheet
DairyNZ. (n.d.). Forage Value Overview. Retrieved January 21, 2015 from DairyNZ:
www.dairynz.co.nz/feed/cultivar-‐selection/about-‐fvi/fvi-‐overview
56
DairyNZ. (2013, April). Technical Series -‐ Issue 15. Retrieved January 28, 2015 from
www.dairynz.co.nz/media/424983/technical_series_april_2013.pdf
Davis, S. (2014, November 25). (A. Pellett, Interviewer) Newstead, New Zealand.
Eastwood, C. (2008). Innovative precision dairy systems: a case study of farmer learning and
technology co-‐development. The University of Melbourne, Land and Food Resources,
Agriculture and Food Systems.
Emily M Gray, R. (2014, March). Productivity in the broadacre and dairy industries.
Agriculture Commodities: March Quarter 2014 .
Karanja, F. (2012, March). Dairy Productivity Growth, Efficiency Change and Technological
Progress in Victoria. Research Paper 2012.5 .
FutureDairy Project. (n.d.). Galleries / Video. Retrieved February 23, 2015 from
www.futuredairy.com.au.
Harty, D. (2014, September 26). (A. Pellett, Interviewer) Causeway, Co Kerry, Ireland.
Herd Navigator. (n.d.). Retrieved January 31, 2015 from http://www.herdnavigator.com
Hille, D. (2014, September 10). (A. Pellett, Interviewer) Bonen, Germany.
Yule, I. (2014). Use Of Remote Sensing For Pasture Management. 17th Precision Agriculture
Symposium Of Australia.
Edwards, J. (2014, April 11). Evaluating rates of technology adoption and milking practices on
New Zealand dairy farms. Animal Production Science .
Kerrisk, K. (2014, August 19). (A. Pellett, Interviewer) Campden, NSW, Australia.
O'Donovan, M. (2014). Pasture Progit Index: New Test Proof 2015. Presentation, Teagasc,
Animal & Grassland Research and Innovation Centre, Moorepark.
Macdonald, G. (2013, April). Measuring residual feed intake in heifer calves. Retrieved
January 28, 2015 from DairyNZ Issue 15 Technical Series:
www.dairynz.co.nz/media/424983/technical_series_april_2013.pdf
57
Trotter, M. (2014). Next-‐Gen Technologies For The Grazing Industries -‐ From Yield Mapping
Pastures To Virtual Fencing Of Livestock. 17th Precision Agriculture Symposium Of
Australasia.
Mourik, J. (2014, September 15). (A. Pellett, Interviewer) Maassluis, Netherlands.
Nation, D. (2015, February 26). (A. Pellett, Interviewer)
Nation, D. (2014, February 6). Dairy Futures CRC Briefing. (A. Pellett, Interviewer) Bundoora,
Victoria, Australia.
O'Connor, B. (2014, August 8). (A. Pellett, Interviewer) Dookie, Victoria, Australia.
O'Donavan, M. (2014, September 22). (A. Pellett, Interviewer) Moorepark, Ireland.
Pedofsky, J. (2014, December 1). Agronomist, Cropmark Seeds. (A. Pellett, Interviewer)
Ashburton, New Zealand.
Romano, R. (2015, February). (A. Pellett, Interviewer)
Roos, S. (2014, September 18). CRV -‐ ovalert system. (A. Pellett, Interviewer) Arnhem,
Netherlands.
Sahlstrom, P. (2014, September 30). (A. Pellett, Interviewer) Tumba, Sweden.
Smit, H. (2014, June 26). Rabobank Big Data Presentation. (A. Pellett, Interviewer) Utrecht,
Netherlands.
Davis, S. (2013, April). Predicting residual feed intake of lactating cows. Retrieved January 28,
2015 from Dairynz: www.dairynz.co.nz/media/424983/technical_series_april_2013.pdf
Tobe, F. (2014, November 12). Are agricultural robots ready? Retrieved February 23, 2015
from robohub: www.robohub.org
Waghorn, K. (2013, April). Efficient calves are efficient cows. Retrieved January 28, 2015 from
Dairynz.co.nz: www.dairynz.co.nz/media/424983/technical_series_april_2013.pdf
Zoetis. (2012). Clarifide, Using net merit for commercial herd success.
58
Plain English Compendium Summary
Project Title: Improving productivity for Australian pasture-based dairy farming.
Nuffield Australia Project No.:
1416
Scholar: Aubrey Pellett Organisation: A O Pellett & J A Morrison
Phone: 61 429667699 Email: [email protected] Objectives
To identify the new technologies, management practices and other factors that are likely to improve dairy farm productivity in the future.
Background
In the past decade productivity growth on Australian dairy farms has slowed. As farm profitability is determined largely by international milk prices, rising input costs and variable seasonal conditions; long term productivity improvement is required to maintain and increase real returns.
Research Research involved travel for in excess of 20 weeks to countries including Australia, New Zealand, United States, United Kingdom, Ireland, Germany, the Netherlands, Sweden, China, Canada and the Philippines. Information was collated from interviews with farmers, academics, commercial product managers, farm consultants, scientists, and review of published and un-‐published papers as well as internet searches.
Outcomes Significant productivity opportunities have been identified in the areas of batch milking robotic systems, paddock robots, genomic assisted ryegrass breeding, pasture selection indexes, development of novel endophytes, growing fodder beet as a high yielding crop, incorporating feed conversion efficiency into a livestock breeding index, using genomic information to guide herd management decisions, mining the data that precision dairy makes available and sharing farm information to facilitate greater supply chain integration.
Implications To promote the adoption of these productivity opportunities the dairy industry could; support development of a new technology finance model, publish fodder beet guidelines for Australian conditions, facilitate a dairy data co-‐op, provide a ryegrass cultivar selection index and set-‐up a new demonstration farm to showcase precision dairy.
